From: Miguel Carrasquer
Message: 32286
Date: 2004-04-24
>Thanks for the links. For those who find it too much effort,I express distaste for the way Spa formulates the rule. The
>wereMiguel dug up four lengthening rules from one source,
>namely "Cours élémentaire de phonétique et de phonologie
>françaises", by one J.J. Spa, Groningen 1978:
>
>1) L'allongement du^ aux consonnes allongeantes
>
>before [v], [z], [Z], [r] and [vr], and provided the vowel is at the
>end of the "groupe rythmique" (syntactic stress), lengthening occurs:
>
>tu ouvres [ty u:vr]
>c'est rare [sE ra:r]
>une voiture neuve [yn vwatyr noe:v]
>
>The [y] in <voiture> is not lengthened, as it is not at the end of
>the
>"groupe rithmique" (G.R.). But "La voiture" by itself is [la
>vwaty:r].
>
>But:
>charte [Sart] (the group [rt] does not lengthen).
>
>2) L'allongement des voyelles nasales
>
>A nasal vowel is lengthened at the end of the G.R. if one or more
>consonants follow:
>
>elle e'tait grande [El etE grA~:d]
>But:
>il e'tait bon [il etE bO~] (no cons. follows)
>bonte' [bO~te] (not at the end of G.R.)
>(My dictionary gives [bO~.te], and generally gives half-long for
>cases
>of lengthening when not at the end of the G.R.)
>
>3) L'allongement compensateur
>
>A vowel is lengthened (at the end of the G.R.) when there *used to
>be*
>an [s] after it (sic!).
>
>fe^te [fE:t]
>ma^le [mA:l]
>mai^tre [mE:tr]
>But:
>mettre [mEtr] (no [s] in Old French)
>
>4) L'allongement de [E] devant consonne non aparie'e (facultatif)
>
>At the end of the G.R., [E] may be lengthened if folowed by a single
>consonant, *provided* the consonant is written (sic) as a single
>consonant and is followed, in writing (sic), by an <e>.
>
>reine [rE(:)n]
>ze`le [zE(:)l]
>But:
>belle [bEl] (written with two <l>s)
>renne [rEn] (written with two <n>s)
>miel [mjEl] (no <e> follows)
>
>Rules (1) and (2) don't seem to be phonemic (unless there are some
>consequent neutralisations). Rule (3) is one most of us are
>acquainted with.
>
>Rule (4) curious, as it seems not to apply before stop consonants,
>but is actually the rule that looks applicable to <mètre>. Miguel
>expresses distaste for it because it is orthographic,
>but I don'tIt's /pA:t/ (except if you don't have /A/ in your French, as
>think that is a problem. This rule seems to relate to the length
>contrast between <jatte> and <Jacques> as well.
>
>What seems to have happened is that besides the major length
>differences caused by the vocalisation of /s/ before consonants
>(Rule 3), vowels in open syllables were longer than vowels in closed
>syllables. When E muet became silent, this lesser length difference
>remained. Under this scenario, degemination would have proceeded in
>parallel with the silencing of E muet.
>
>The limited data I have is:
>/a/: <patte>, <jatte>
>/a./: <limonade>
>/a:/: <Jacques>, <pâte>
>/e/: <mettre>, <belle>, <renne>, <miel>
>/e./: <mètre>, <reine>, <zèle>
>/i/: <riz>, <bis>
>/i:/: <rire>, <mobile>
>
>I don't have an explanation for the contrast of /a./ and /a:/. I
>suspect the gemination in <Jacques> is purely orthographic. I am
>puzzled as to why <pâte> is /pa:t/ and not /pA:t/.
>For <métre>, IIt's of course a learnèd word, from the Greek, and
>assume the syllabification was me-tre rather than met-re.
> chèvre, not cap-ra > *chav-re).=======================